---
title: "Product Information"
output: rmarkdown::html_vignette
vignette: >
  %\VignetteIndexEntry{Product Information}
  %\VignetteEngine{knitr::rmarkdown}
  %\VignetteEncoding{UTF-8}
---

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)
eval_chunks <- curl::has_internet()
```

Before you start downloading data, you first need to know which
data you need and for which purpose. And not unimportant: you need to know
where to find it. Although data discovery is not the primary objective of
this package, it provides several instruments to obtain product information.

## The Web Browser

Perhaps the easiest way to find products and layers is via the
[online catalogue](https://data.marine.copernicus.eu/products). To
streamline your workflow using the web browser, you should check
`vignette("translate")`. That vignette explains how you can copy a
request from the web browser catalogue and use it in R.

## Listing Products

To get a complete overview of all available products, you can list them with
either `cms_products_list()` or `cms_products_list2()`:

```{r listing-products, message=FALSE, eval=eval_chunks}
library(CopernicusMarine)

cms_products_list() |> head(3)

cms_products_list2() |> head(3)
```

The first returns a `data.frame` complete with al sorts of meta information
about the type of variables in the product, the spatio-temporal coverage,
number of vertical layers, etc. You can use this `data.frame` to narrow your
search by applying a `dplyr::filter()` on it.

You can even pass arguments that are used to search the online catalogue.
These are not well documented. The example below shows how to search for
free text and filter on area and variables:

```{r searching-products, message=FALSE, eval=eval_chunks}
cms_products_list(freeText = "wave",
                  facetValues = list(areas             = list("Europe"),
                                     specificVariables = list("Velocity")))
```

The reason this is poorly documented is because this function does not
use the formal API. Instead it uses the web-form used by the online catalogue.
Users should therefore not rely on it too much as it may get discontinued or
altered at any time.

Instead, users can use `cms_products_list2()` which produces a list of products,
by using the official API. Unfortunately, this list does not contain any additional
information. For this purpose users can refer to `cms_product_metadata()`,
`cms_product_details()`, or `cms_product_services()` as described below.

## Product Details

With `cms_product_details()` you will get some descriptive information about your
product. Nothing too fancy, but it will help you understand what the product
is all about.

```{r product-details, eval=eval_chunks}
cms_product_details("GLOBAL_ANALYSISFORECAST_PHY_001_024") |> summary()
```

## Product Metadata

When subsetting  a dataset with `cms_download_subset()`, the most
tricky thing is discovering what the available ranges are for its dimensions.
You can use `cms_product_metadata()` for this purpose. It returns a named list.

```{r product-metadata, eval=eval_chunks}
meta_info <- cms_product_metadata("GLOBAL_ANALYSISFORECAST_PHY_001_024")

## Get the dimension properties for the first layer in this product
meta_info$properties[[1]]$`cube:dimensions` |> summary()
## Get the variable properties for the first layer in this product
meta_info$properties[[1]]$`cube:variables` |> summary()
```

Another way to get the dimension ranges is by setting up a stars proxy
object (see `vignette("proxy")`) and call `st_dimensions()` on it:

```{r product-proxy, warning=FALSE, eval=eval_chunks}
library(stars) |> suppressMessages()

myproxy <- cms_zarr_proxy(
    product       = "GLOBAL_ANALYSISFORECAST_PHY_001_024",
    layer         = "cmems_mod_glo_phy-cur_anfc_0.083deg_P1D-m",
    asset         = "timeChunked")
st_dimensions(myproxy)
```

## Product Services

If you want a more raw access point to your data, you can use `cms_product_services()`.
It will present a `data.frame` for your product with all services provided by
Copernicus. You can use any of the columns with the `"_href"` suffix to get an
URL of a specific service. If you want to access those directly, you are on your
own. It is easier to use any of the wrappers provided by this package to access
the data.

```{r services, eval=eval_chunks}
cms_product_services("GLOBAL_ANALYSISFORECAST_PHY_001_024")
```